gsskrb5: CVE-2022-3437 Avoid undefined behaviour in _gssapi_verify_pad()

By decrementing 'pad' only when we know it's safe, we ensure we can't
stray backwards past the start of a buffer, which would be undefined
behaviour.

In the previous version of the loop, 'i' is the number of bytes left to
check, and 'pad' is the current byte we're checking. 'pad' was
decremented at the end of each loop iteration. If 'i' was 1 (so we
checked the final byte), 'pad' could potentially be pointing to the
first byte of the input buffer, and the decrement would put it one
byte behind the buffer.

That would be undefined behaviour.

The patch changes it so that 'pad' is the byte we previously checked,
which allows us to ensure that we only decrement it when we know we
have a byte to check.

Samba BUG: https://bugzilla.samba.org/show_bug.cgi?id=15134

Signed-off-by: Joseph Sutton <josephsutton@catalyst.net.nz>
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

lib/gssapi/krb5/decapsulate.c

@@ -193,13 +193,13 @@ _gssapi_verify_pad(gss_buffer_t wrapped_token,
     if (wrapped_token->length < 1)
 	return GSS_S_BAD_MECH;
 
-    pad = (u_char *)wrapped_token->value + wrapped_token->length - 1;
-    padlength = *pad;
+    pad = (u_char *)wrapped_token->value + wrapped_token->length;
+    padlength = pad[-1];
 
     if (padlength > datalen)
 	return GSS_S_BAD_MECH;
 
-    for (i = padlength; i > 0 && *pad == padlength; i--, pad--)
+    for (i = padlength; i > 0 && *--pad == padlength; i--)
 	;
     if (i != 0)
 	return GSS_S_BAD_MIC;